Re: Optimisation functions missing? (fwd)

[Jonathan King]

I immediately forwarded Heber Farnsworth's message to an accomplished
signal processing person I know, and got the following answer back
(with permission to re-forward).  Maybe it's not the answer you wanted
to hear, but the suggested people to contact might be helpful.

jking

---------- Forwarded message ----------
Date: Thu, 4 Mar 1999 11:56:58 -0800 (PST)
From: address@hidden
To: address@hidden
Subject: Re: Optimisation functions missing? (fwd)

Jon,

you can forward this to the group if you want. They should not 
disregard the simplex method even if they implement an interior point
method. Simplex is faster for some types of problems.

They should consult Gill and his notes for a graduate course that he
teaches. Gill is in the math department at UCSD. He is more than willing
to share his notes.

Solving constrained optimization is the hard problem in this. Picking a
paper in the field and implementing a code based on that is not the thing
to do. There are now hundreds of thousands of algorithms for interior
point mehtods. Which are numerically better is still an open question.

It is so much a developing field that I would not advice anyone myself
because I am not familiar with the latest develoments. Gill is one good
person to approach. Yinyu Ye at U of Iowa is another. That's a start.

Best of luck,
Irina

[copy of forwarded message from address@hidden deleted by jking]